Dynamic Peak-Load Pricing ∗

This paper develops a dynamic theory of peak-load pricing and service provision. Unlike the literature on peak-load pricing which assumes that the demand for service in each period is predetermined, here, consumers are utility maximizers and are free to postpone buying the service to a different season. In turn, service providers can manipulate the price system to shuffle consumers between seasons as to maximize profit or social welfare. Results show the conditions under which the classical peak-load pricing algorithm may need to be modified since a firm or a social planner will find less beneficial to delay the service when the time discount factor is taken into account. The paper also shows that peak-off-peak price- or quantity-reversals cannot be realized in an environment where consumers are allowed to postpone buying the service.